Broad Match Keywords Are Google’s Default. Here Is What That Costs You.

Nobody chooses broad match. That’s the part most business owners miss.

You type a keyword into Google Ads. You don’t add quotation marks, because nobody told you to. That keyword is now broad match, and it’ll show your ad on searches that share no words with what you typed.

Google calls this reach. Your bank statement calls it something else.

We audit a lot of accounts. In most of them, somewhere between a quarter and a half of the search budget went to searches the owner would have rejected on sight. Not fraud. Not a bug. Just the default setting doing exactly what it was designed to do.

This post shows you what broad match keywords actually do to a small budget, how to measure the damage in your own account, and how to contain it without losing the customers you do want.

What Broad Match Actually Does

Broad match lets Google show your ad on any search it judges to be related to your keyword.

Read that again. Not searches containing your keyword. Searches related to it.

Google’s own documentation gives the example openly. Bid on “luxury watches” and your ad can appear for “high-end timepieces” or “expensive wristwatches”. No shared words. Google decided the meaning was close enough.

That judgement is made by an algorithm, in a fraction of a second, on a search you’ll never see until you go looking for it.

Two of those matches will be excellent. The third will be a student writing an essay.

Here’s the bit that catches people out. Broad match doesn’t just add searches around the edges of your keyword. It replaces your judgement with Google’s on every single auction.

You picked the keyword. Google picks the customer.

The Three Words That Should Worry You

The whole problem sits in one phrase Google uses: “related to your keyword”.

Related isn’t a fixed distance. It moves. It has moved a lot since 2021, when Google folded broad match modifier into phrase match and started leaning on language models to interpret meaning.

In practice, related now covers a wider circle than it did three years ago. That circle includes people researching, people comparing, people looking for jobs, and people who want the thing you sell but can’t afford it.

Every one of them costs you the same as a buyer.

The Change In 2021 That Most Owners Never Heard About

If your account is more than a few years old, something happened underneath it that nobody emailed you about.

Google used to offer a fourth option called broad match modifier. You put a plus sign in front of the words that had to appear, and Google had to honour it. It gave you reach with a handrail.

In 2021, Google retired it and folded its behaviour into phrase match. At the same time, the company began leaning harder on language models to interpret what a search means rather than what it says.

The practical effect was that every match type loosened by one notch. Phrase match started behaving a bit like the old modifier. Broad match started behaving like nothing that had existed before.

So if you set your campaigns up in 2019 and haven’t touched the keywords since, they aren’t doing what you told them to do. The words are the same. The rules underneath changed.

This matters most for the accounts nobody has looked at in a while, which in our experience is most of them.

Why Google Switched It On For You

This is worth saying plainly, because it explains everything else.

Google makes money when your ad shows and gets clicked. Broad match makes your ad show on more searches. There’s no version of this where the default serves your budget before it serves theirs.

That doesn’t make broad match a scam. It makes it a setting with an owner, and the owner isn’t you.

Google’s guidance is honest about the conditions. Their documentation says broad match should be paired with Smart Bidding, because bids need to reflect the signals of each search. That’s a real prerequisite, quietly stated.

Most small accounts run broad match without meeting it.

What The Waste Actually Looks Like

Waste isn’t abstract. It has a shape, and after a few hundred audits the same six patterns come up again and again.

Job seekers. A plumbing firm bids on “emergency plumber”. The account pays for “plumber jobs near me”, “plumber apprenticeship”, “how much do plumbers earn”. Nobody in that group is buying a service.
Free and DIY. A cleaning company bids on “oven cleaning”. It pays for “how to clean an oven with baking soda”. That person has already decided not to hire anyone.
Competitor research. You bid on your service. You pay for people typing a rival’s brand name plus the word “reviews”. They’re not shopping. They’re checking.
Wrong service. A roofer bidding on “flat roof repair” pays for “flat roof insulation grant”. Same trade, different job, no revenue.
Wrong geography. A firm covering Bedfordshire pays for searches from a city three hundred miles away, because the search itself carried no location and Google matched on meaning alone.
Wrong stage entirely. “What does a new boiler cost” and “boiler installation near me” are not the same person. One is nine months from buying. The other wants a phone number.

In one documented B2B audit, a single core product term was being matched to more than 1,400 irrelevant queries, including job seekers, free-tool hunters and out-of-market geographies.

That’s the same pattern a local business sees. Fewer zeros, identical shape.

The Numbers Nobody Gives You

Here’s the benchmark that lets you score your own account.

A search account running on tight match types, with a maintained negative list, usually shows an irrelevance rate under 10 per cent. That’s normal. No account is perfect.

A broad match heavy account commonly sits between 30 and 60 per cent.

Anything above 25 per cent isn’t bad luck. It points to a structural mismatch between what you bid on and what Google matched you to.

There’s also a large study worth knowing. Optmyzr looked at roughly 2,600 accounts and found that in 85.6 per cent of them, click-through rate was higher on exact match than on broad.

Read what that means. On the vast majority of accounts, people were less inclined to click the broad match ads. They looked less relevant because they were less relevant.

Broad match usually wins on cost per click and loses on cost per customer. Those are different races, and only one of them pays your wages.

A Worked Example, Line By Line

Abstract percentages are easy to nod at and easy to ignore. So here’s the same problem with the numbers filled in.

Take a local trade business spending £900 a month on search. Average cost per click £3.60. That’s 250 clicks a month.

The account has four keywords, all broad match, no negative list. Standard setup for a business that built its own campaigns in an afternoon.

Now sort the search terms report by cost and group what’s there.

Genuine service searches: 132 clicks, £475. These are the customers.
Job and salary searches: 41 clicks, £148. Nobody here is hiring you.
How-to and DIY searches: 38 clicks, £137. They’ve already decided to do it themselves.
Competitor brand searches: 22 clicks, £79. Window shoppers checking a rival.
Out-of-area searches: 17 clicks, £61. Real intent, wrong county.

That’s £425 a month, 47 per cent of the budget, going to people who could never have become customers.

At a 20 per cent enquiry rate on genuine traffic, those 132 real clicks produce about 26 enquiries. The other 118 clicks produce nothing.

Recover even half of that £425 and put it back into genuine searches. You buy roughly 59 more real clicks and pick up around 12 more enquiries a month, on the same budget.

No new spend. No new campaign. Same ads, same landing page. The only thing that changed is who saw them.

That’s the argument for reading your search terms report, written in money.

Why A Small Budget Suffers Most

This is the part page one never explains, and it’s the reason broad match hurts a local business more than a national one.

Write out the arithmetic.

Say you spend £30 a day. Your average cost per click is £3. That buys you ten clicks. One in five enquiries turns into a job.

If every click is a real prospect, you get two enquiries a day. Ten a week.

Now let 40 per cent of that traffic go to job seekers and DIY searches. You still buy ten clicks. Only six are prospects. You get 1.2 enquiries a day.

Same spend. Six enquiries a week instead of ten. Nothing on the dashboard turned red.

A business spending £3,000 a day absorbs that. It buys a thousand clicks, and the algorithm has enough conversions to learn from even after the waste is stripped out.

You don’t have that cushion. At ten clicks a day, waste isn’t a rounding error. It’s most of your week.

There’s a second cost, and it’s worse. Smart Bidding learns from what converts. Feed it four hundred junk clicks a month and you haven’t just lost the money. You have taught the system a slightly wrong idea of who your customer is.

Why It Stays Silent

Broad match waste doesn’t announce itself. That’s the whole problem.

An account with 40 per cent waste doesn’t throw an error. Impressions look healthy. Clicks look healthy. Your cost per click may even have fallen, which reads as good news.

The only place it shows up is cost per enquiry, and most owners have nothing to compare that number to. Is £84 per enquiry bad? Without a baseline, it’s just a number on a screen.

So the account keeps running. The monthly invoice gets paid. Nobody looks at the search terms, because nobody knows the report exists.

That’s why we call it silent. It never fails loudly enough to get investigated.

How To Check Your Own Account In Ten Minutes

You don’t need a tool or an agency for this. You need one report and ten quiet minutes.

Open Google Ads and pick your search campaign.
Go to Insights and reports, then Search terms.
Set the date range to the last 30 days.
Sort by Cost, highest first. Not by clicks, not by impressions. Cost is the only column that spends your money.
Read the top 50 rows and ask one question of each: would I have chosen to pay for this search?

Keep a running total of the spend on every row where the answer is no.

Divide that total by your total spend for the period. That percentage is your irrelevance rate.

Under 10 per cent, you’re fine. Between 10 and 25 per cent, tighten up. Over 25 per cent, broad match is costing you real money every week.

Do this once, and you’ll never guess about your account again.

What That Report Won’t Show You

One honest caveat, because you’ll notice the gap eventually and wonder if you’ve done something wrong.

The search terms report doesn’t show you everything. Google withholds searches made by only a handful of people, on privacy grounds.

So the numbers won’t reconcile. Add up the clicks in your search terms report, and you’ll usually find they fall short of the clicks in your campaign totals. Sometimes by a little. Sometimes by a third.

That missing slice isn’t a rounding error, and it isn’t visible. It’s the long tail, and on a broad match campaign the long tail is exactly where the strangest matches live.

Two things follow from that.

First, your irrelevance rate is a floor, not a ceiling. The real figure is probably worse than what you can see.

Second, this is another argument for pattern-based negatives. You can’t block a search you’ll never be shown. You can block the word it almost certainly contains.

Nobody at Google will tell you this. It’s worth knowing before you spend an afternoon trying to make the columns add up.

Fix One: Find Out What Match Type You Are Actually On

Most owners genuinely don’t know. Check before you change anything.

Open the Keywords tab and look at how each keyword is written. Plain text is broad match. Quotation marks mean phrase match. Square brackets mean exact match.

Then check campaign settings for the broad match keywords toggle. Google added a switch that converts every keyword in a campaign to broad match in one click. It’s easy to enable by accident and easy to miss.

If that toggle is on and you did not turn it on, that’s your answer.

Fix Two: Build The Negative List You Should Have Had At Launch

Negative keywords tell Google what you never want to pay for. Most small accounts have none.

Start with the universal blockers. Free, cheap, DIY, how to, jobs, salary, vacancy, career, course, training, apprenticeship, wholesale, second hand, template, meaning, definition.

Then add your own trade’s traps. A roofer blocks “grant” and “insulation”. An accountant blocks “software” and “jobs”. A solicitor blocks “legal aid” if they don’t offer it.

Add these at account level using a shared negative keyword list, not campaign by campaign. Build it once, apply it everywhere, and every new campaign inherits your protection on day one.

One caution. Negatives only stop searches you have already paid for and thought to block. They’re a containment tool, not a cure.

Fix Three: Block Patterns, Not Single Searches

This is the mistake that keeps people busy and poor.

An owner opens the search terms report, sees “plumber jobs in Luton”, adds it as an exact negative, and feels productive. Next week the same money goes to “plumber jobs in Dunstable”.

Don’t block the search. Block the pattern.

Add “jobs” as a phrase match negative, and you kill every variation at once. One entry, permanent effect.

The rule of thumb: if you can imagine five versions of the same bad search, you’re looking at a pattern. Find the one word they share and block that instead.

Phrase match negatives do most of the work. Keep exact match negatives for the rare case where one specific search is bad but close relatives are fine.

Fix Four: Put Broad Match In A Cage

If you want to keep broad match, don’t let it share a budget with your proven keywords.

Give it its own campaign with its own capped budget. Ten or twenty per cent of your total, no more.

Now broad match spend is visible and contained. You can see exactly what it costs and exactly what it returns, and it can’t quietly eat the budget that was funding your best exact match terms.

Your phrase and exact campaigns carry the reliable volume. The broad campaign becomes a controlled experiment for finding new search terms you had not thought of.

That’s broad match used as a research tool. That’s the version that earns its place.

Fix Five: Give The Algorithm Something Real To Learn From

Broad match plus Smart Bidding only works when the bidding half has real data.

Check what your account counts as a conversion. If it counts page views, or clicks on a phone number that nobody answered, or a form fill from a job applicant, then the algorithm is optimising towards the wrong person.

Count enquiries that could become revenue. Nothing else.

Get that right and broad match becomes far less dangerous, because the system is finally chasing the correct target. Get it wrong and broad match multiplies the error across every search Google decides is related.

Fix the conversions first. Then reconsider the match types.

When Broad Match Is Actually The Right Call

This isn’t an argument for never using it. Broad match has a real job.

Use it when your exact match keywords are starved of volume. Some niches are genuinely small, and a tightly matched account can sit there getting forty impressions a week. Broad match finds demand you couldn’t have guessed the wording of.

Use it when you’re new to a market and don’t know how customers describe the problem yet. The search terms report from a caged broad campaign is the cheapest market research you’ll ever buy.

Use it when you have the prerequisites in place. All of them, not most.

Here’s the checklist. Broad match is safe when you have Smart Bidding running on genuine conversion data, at least thirty real conversions a month to learn from, a maintained negative keyword list, someone reading the search terms report weekly, and enough budget that a learning period won’t sink the month.

Score yourself honestly. Most small accounts meet two of the five.

If you meet two, you’re not ready. Run phrase and exact, build the data, and revisit it in three months.

What Good Looks Like After Sixty Days

Set expectations properly, because the first fortnight can look worse before it looks better.

Weeks one and two: your impressions fall. This is correct. You have stopped showing your ad to people who were never going to buy.

Weeks three and four: cost per click often rises. Also correct. You’re now competing for searches that other real businesses want, and those cost more.

Weeks five to eight: cost per enquiry starts to fall. Your click volume is lower, and your enquiry volume is level or up.

That’s the trade. Fewer clicks, better clicks, more of your budget landing in front of people who can actually buy.

If you only watch impressions and clicks, this looks like a downgrade. Watch cost per enquiry instead, and it looks like what it’s.

The Mistakes We See Most

A short list, drawn from accounts we have taken over this year.

Adding negatives once at launch and never again. The list is a habit, not a task.
Blocking single searches instead of patterns. Endless work, no lasting effect.
Judging match types on cost per click. Cheap clicks that never convert are the most expensive thing in the account.
Turning broad match off everywhere in a panic. You lose the discovery as well as the waste. Cage it, don’t kill it.
Adding a negative that’s too wide. Block “cheap”, and you might block “cheapest emergency plumber”, which is a real customer in a hurry. Check before you add.
Leaving the campaign-level broad match toggle on while carefully writing phrase match keywords. The toggle overrides your work.

Your Thirty-Minute Job This Week

One sitting. No tools. Do it on a Tuesday morning while the phones are quiet.

Minutes one to ten: run the search terms report as described above and work out your irrelevance rate. Write the number down.

Minutes eleven to twenty: build a shared negative keyword list. Add the universal blockers, add your trade’s traps, add every pattern you spotted in the report. Apply it to all search campaigns.

Minutes twenty-one to twenty-five: check your keyword match types and the campaign-level broad match toggle. Move your five best-performing keywords to phrase match if they’re currently broad.

Minutes twenty-six to thirty: open your conversion actions and confirm you’re counting enquiries, not page views.

Put a repeating fifteen-minute slot in the diary for the same time next week. That single habit does more for a small account than any bid strategy change you’ll ever make.

Frequently Asked Questions

Should I use broad match keywords?

Only if you have Smart Bidding running on real conversion data, a negative keyword list already built, and enough budget to absorb a learning period. If any of those three are missing, start on phrase and exact match. You can always widen later. Getting the money back is harder.

How do I know if broad match is wasting my budget?

Open your search terms report for the last 30 days, sort by cost, and read the top 50 rows. Add up the spend on searches you would never have chosen to pay for. If that figure is more than 25 per cent of your total spend, broad match is costing you money every week.

How do I turn off broad match in Google Ads?

There’s no single off switch at account level. Match type is set on each keyword. Edit the keyword and wrap it in quotation marks for phrase match, or square brackets for exact match. Also check your campaign settings for the broad match keywords toggle, which switches every keyword in that campaign to broad in one click.

Do negative keywords fix broad match?

They contain it rather than fix it. A negative only blocks a search after you have paid for it once and thought to add it. Treat the list as a weekly habit, and it works well. Treat it as a one-off setup task, and the waste creeps back within a month.

Is broad match cheaper than exact match?

The cost per click is usually lower, because there is less competition for loosely related searches. The cost per customer is usually higher, because more of those cheap clicks had no buying intent. Judge every match type decision on cost per enquiry, never on cost per click.

How much of a Google Ads budget is typically wasted?

A tightly matched account with a maintained negative list usually runs under 10 per cent irrelevant spend. Broad match heavy accounts commonly sit between 30 and 60 per cent. Anything over 25 per cent points at a structural problem you can fix, not bad luck you have to accept.

Will my leads drop if I move away from broad match?

Your clicks will drop. Your leads usually hold steady or rise, because the budget stops funding people who were never going to call. Give it six weeks and judge on enquiries, not traffic.

Does broad match work better now that Google uses AI?

It understands meaning better than it did in 2021, and it makes fewer nonsense matches. It has also widened what counts as related. Better matching within a wider circle isn’t the same as safer matching, and the prerequisites still apply.

Start With The Report, Not The Settings

Broad match isn’t the villain. Broad match running without supervision, on a small budget, with no negative list and no conversion data, is the villain.

The good news is that this is one of the few problems in Google Ads you can measure yourself, this week, for free.

Open the search terms report. Sort by cost. Read fifty rows and be honest about what you see.

Whatever that percentage turns out to be, you’ll know something about your account that you did not know this morning. Most of your competitors still won’t.

That’s where every good account starts. Not with a clever bid strategy, but with an owner who finally looked at what they were paying for.

If you want a second pair of eyes on the numbers, that’s exactly what we do at SF Digital. Either way, run the report.

Did You Enjoy This Blog Post?

I hope you enjoyed this blog post, and thank you so much for being here. We also upload videos to our YouTube channel every weekday. Please subscribe so you are one of the first to be notified.

If you enjoyed this blog, you may also like:

SF Digital Studios

Search This Blog